Immunohistochemistry on Fluid Specimens: Technical Considerations Blake Gilks Dept of Pathology University of British Columbia, Vancouver, BC, Canada
Disclosures None
Learning Objectives At the end of this presentation, participants will: 1. be aware of strategies for handling immunostaining on non-formalin fixed/paraffin embedded (FFPE) specimens 2. be able to describe some of the problems encountered when staining protocols optimized for use on FFPE samples are applied, unchanged, to specimens fixed in CytoLyt 3. have a plan for handling requests for class II immunostains on specimens other than FFPE tissue blocks
Limitations of our menu of Abs Initially validated for FFPE samples only Expansion of validation/controls to include bone marrow biopsies Careful attention to pre-analytical variables (neutral buffered formalin, adequate fixation) Cytology specimens?
Liquid-based cytology specimens Popular Good diagnostic yield (at least in era of benign vs malignant distinction Increasing need for subtype diagnosis for management What to use as controls for immunostaining?
Background on handling of cytology specimens at Vancouver General Hospital Peritoneal and pleural fluids specimens: fixed in CytoLyt -> thin prep cytology and cell block FNA: collected in CytoLyt -> cytospin and cell block
Pragmatic Considerations for Fluid Specimens Some recurrent differential diagnoses (e.g. adenocarcinoma vs mesothelioma) so a limited panel of Abs used Often accompanying or subsequent surgical pathology specimens, so direct comparisons are possible
Approach to IHC on CytoLyt fixed specimens Observational study (often accompanying surgical pathology specimen Formal comparison
Observational Study Design Approach Dr. A. Churg Ask him to evaluate performance of Abs when applied to pleural or peritoneal fluid specimens and report back
Observational versus Formal Validation OBSERVATIONAL Pros easy inexpensive (?) traditional approach Cons inconclusive operator specific FORMAL VALIDATION Pros easily interpreted data generalizable results Cons impossible to validate full IHC menu
Antibodies That Appear to Work Properly With the Same Conditions Used for FFPE Tissue Pan-keratin CK7 CK5 CK20 CEA MOC-31 TTF-1 BerEP4 Cdx2 Naspsin p63 HepPar1 (HSA) CD34 ER PR Synaptophysin Chromogranin CD56 Calcitonin CD3 CD20 CD45 Smooth muscle actin DOG1 c-kit CD10
Antibodies That Produce Problems with CytoLyt CD138: Does not stain (probably does if conditions are changed) S-100: Stains some adenocarcinomas WT-1: Does not stain benign mesothelial cells in fluids at the same concentration as used for FFPE tissues However it does stain most ovarian serous neoplasms Calretinin: Does not stain benign mesothelial cells in fluids at the same concentration/conditions as used for FFPE tissues Her-2: Appears to be less sensitive than in FFPE
Identification of Mesothelial Cells in Fluids Problem: Calretinin and WT-1 at the concentrations that work on mesothelial cells in formalin fixed tissue don t stain mesothelial cells in CytoLyt fixed fluids Possible solution: Increase antibody concentrations and change conditions re heat or no heat retrieval
FFPE and CytoLyt Staining Conditions Calretinin (CellMarque) FFPE: 1:250 with heat retrieval CytoLyt: 1:100 no heat WT-1 (Dako) FFPE: 1:100 with heat retrieval CytoLyt: 1:25 with heat retrieval
VC12-3921 Known CLL. Pericardial Effusion
Calretinin
WT-1
VC12-3888 Pericardial Effusion Patient with Dilated Cardiomyopathy
Calretinin-Notionally Optimized Conditions
WT-1
CK5
Tentative Conclusions Changing conditions and antigen retrieval improves recognition of mesothelial cells, but most cells that are probably mesothelial still don t stain for Calretinin or WT-1 in CytoLyt fixed material Calretinin staining is particularly inconsistent Unclear if further increasing antibody concentrations will fix the problem Combination of WT-1 and CK5 may be useful for identifying mesothelial cells
Staining of Ovarian Serous Carcinomas WT-1 usually works with the same conditions used for FFPE ER works but is often weaker than in FFPE
Metastatic serous ca in pleural fluid VC12-3902
WT-1 (FFPE conditions)
ER (FFPE conditions)
MOC-31 (FFPE conditions)
False Staining for S-100 in CytoLyt Fixed Material Some adenocarcinomas that otherwise mark as adenocarcinomas stain for S-100 in CytoLyt fixed material
VC12-3956 Known Adenoca of Lung Pleural Fluid
TTF-1
S-100 CytoLyt Fixed
S-100 Not CytoLyt Fixed
Formal Evaluation Choose a specific clinical scenario, state a hypothesis, design a conventional experiment Coerce resident into doing the work
Background: High-grade serous carcinoma Usual presentation: ascites, intraperitoneal spread. Neo-adjuvant chemotherapy: less morbidity and identical patient survival. Diagnoses established on ascitic fluid or omental core biopsy. Exclude other tumours (clear cell carcinoma) that are resistant to platinum-based chemotherapy. Vergote I et al. N. Engl. J. Med. 2010 Sep 2;363(10):943-953
Background: Immunohistochemistry. WT-1 and ER are markers for serous carcinomas (vs other ovarian subtypes, GI or GU carcinomas, or mesotheliomas). Study to compare IHC expression from cytological preparations and matched surgical specimens.
Material and Methods Histologically confirmed cases of serous carcinoma (ovary, fallopian tube, peritoneum) with positive ascitic fluid/ peritoneal washings. Cell block preparation: 1. Received fresh, centrifuged 5 min 1200G, supernatant removed. 2. ThinPrep CytoLyt Solution (Hologic Inc), centrifuged 5 min 1200G, supernatant removed 3. HistoGel (Thermo Scientific) 4 o C for 5 min. 4. 10% Formalin for minimum of 24hrs.
Material and Methods Surgical blocks: Fixed in 10% formalin within 3 hours of removal for minimum of 24 hours. IHC performed with automated Ventana systems protocol including heat antigen retrieval. ER (Labvision, SP1, 1:50). WT-1 (DAKO, 6F-H2, 1:100).
Results 26 cases: 23 high, 3 low-grade serous carcinoma. Ave age at surgery 68 (44-92) years old. Tumour cells not identified in re-cut from one surgical specimen + insufficient number of tumour cells in one section re-cut from cell block. Statistical analysis performed on remaining 25 (WT-1) and 24 (ER) cases.
Results: Zonal pattern of staining WT-1 (A) and ER (C) staining at well fixed/ peripheral portions of the specimen versus WT-1 (B) and ER (D) in poorly fixed/ central portions of specimen. For our study: assessment of expression performed on well fixed portions only.
Results: high-grade serous carcinoma Surgical specimens: H+E (A), WT-1 (B) and ER (D). Cytological preparations: WT-1 (C) and ER (E).
Results: low-grade serous carcinoma Surgical specimens: H+E (A), WT-1 (B) and ER (D). Cytological preparations: WT-1 (C) and ER (E).
Results WT-1 expression: Surg +ve Surg -ve Cyto +ve 20 1 Cyto -ve 4 0 ER expression: Surg +ve Surg -ve Cyto +ve 16 0 Cyto -ve 7 1 WT-1: +ve in 96% (24/25) of surgical specimens, 84% (21/25) of matched cytological preparations. Difference not statistically significant (p=0.35). ER: +ve in 96% (23/24) of surgical specimens and 66%, (16/24 ) of matched cytological preparations. Statistically significant difference (p=0.02).
Results IHC following neo-adjuvant chemotherapy H+E (A), WT-1 (B), ER (C). Prior chemotherapy did not alter expression of ER and WT-1, as previously described. Miller K et al. J. Clin. Pathol. 2008 May;61(5):652-657.
Discussion Ovarian surface epithelial carcinoma are distinct diseases. Specific recommendations for different subtypes. Platinum-based chemotherapy: 80% of high-grade serous carcinomas respond. Clear and mucinous ca resistant to platinum based therapy. PARP inhibitors for high grade serous carcinoma. Sunitinib and radiotherapy for clear cell carcinoma. HER2: potential target for mucinous carcinoma.
Discussion Diagnoses established on ascitic fluid or omental core biopsy. No evidence that accurate subtype diagnosis is possible based on a cytological specimen. Breast carcinoma studies comparing alcohol-fixed liquid based cytology preparations to surgical specimens show good correlation for ER (80-90% positive agreement). In our study: 70% positive agreement for ER, 83% positive agreement for WT-1. Hanley KZ et al. Cancer 2009 Aug 25;117(4):279-288 Williams SL et al. Int. J. Clin. Exp. Pathol. 2009;2(5):476-480
Discussion WT-1 expressed in 96% (24/25) of surgical specimens and 84% (21/25) of matched cytological preparations. ER expressed in 96% (23/24) of surgical specimens and 66% (16/24) of cytological preparations. Discordant staining (absent staining in cytological preparations when there was positivity in matched surgical specimens) was observed in 17% (4/24) and 30% (7/23) of cases.
Conclusion Useful information can be obtained based on alcohol-fixed liquid-based cytology specimen. Results must be interpreted with caution, as false negative results can occur.
MCQ Which of the following statements about fixation of fluid cytology specimens in alternative fixatives is true? 1. Staining sensitivity is consistently decreased 2. Staining specificity is consistently decreased 3. There are not consistent or predictable changes in staining as a result of using an alternative fixative 4. There are no changes in staining as a result of using an alternative fixative
Acknowledgements Drs. Andrew Churg and Ananta Gurung Margaret Luk, Bev Wolber and Carmen Michelsen